Colibrys is one of the major players in the accelerometer market and has recently released its latest VS1000 accelerometer, targeting the crucial up-and-coming industrial applications. One of the most important reports Yole Développement has launched in 2015 is entitled “High-End Gyroscopes, Accelerometers and IMUs for Defense, Aerospace & Industrial applications”. It analyses the market and applications for high-end inertial systems. In particular, important trends have suddenly emerged in industrial applications and, according to Yole Développement, these applications could play a significant role in the future, potentially reaching $159M in 2019.
Industrial applications could interest a large number of high-end players such as Honeywell and Northrop Grumman, but also Colibrys, which is part of the Safran Group. Yole Développement has the great opportunity of interviewing Mr Patrick Gougeon, CEO of Colibrys, to take a deeper look into his company’s experience and know-how (www.colibrys.com)
Yole Développement: Can you please introduce Colibrys ?
Patrick Gougeon: Colibrys is a Swiss MEMS company, designing, manufacturing and selling the highest standard of MEMS, performance and quality wise. Over our 25 years of experience – and more than 8 million MEMS delivered so far, we have built an undisputed reputation in the most demanding sensor markets – aerospace, military, transportation and energy. We belong to the Safran Group, an international high-technology group of 70,000 employees, and this gives us access to capabilities, skills and management discipline, which fuels our growth through leadership in technology and quality of service for the long term. Our mission is to continuously open new markets for high performance MEMS, by replacing legacy technologies.
YD: Your company has recently been integrated into the Safran group. What has changed in Colibrys with this integration at the company and customer interaction levels ?
PG: Lots of things! Before, the company still relied on the core assets that made it one of the leaders of the high-end sensor market segment: incredible people with a unique set of skills, one of the world’s richest environments in micro-technology with a business culture fostering cross-fertilization, a well proven highly accurate accelerometer MEMS platform that has been improved over 25 years and 1,000,000 sensors manufactured in Switzerland.
What we gain is the power of a large group, whose business is focused in critical technologies for the aerospace and defense market. One characteristic of this group is its focus on core technologies, in particular inertial sensors, where it counts among the best worldwide. Safran is definitely a technology group with a performance leadership position, and therefore we have a lot in common in our DNA.
Fundamentally, this group brings us what we were missing:
- Funding for accelerating our product renewal and to reach unmet performance levels,
- Strict requirements, common to the aerospace and defense sector, that pulls us to a higher level in terms of standards of operational excellence,
- Complementary expertise, as we are increasingly teaming for research and technology activities,
- Ambitions outside the accelerometer space, so that we can develop in the future into a broader family of very high-end sensors.
Apart from what the group is giving us, we made a lot of improvements by ourselves:
- We moved to a brand new fab, fully operational since July 2014,
- Management and sales and marketing has been strengthened.
We have now our own sales structure in place in the USA and China and are working on growing it in other regions.
Our customers value the fact that we are now pushing new products with amazing performance and that the continuity of supply and excellence of service is guaranteed as a standard. Safran would not accept less from us.
We see the future as brighter than ever before.
YD: What makes your products different from others’?
PG: We aim at being the best in long term accuracy in the harshest operational conditions. In fact, our target is to meet the performance currently achieved by other technologies – mechanical and quartz technologies – for demanding customers, with a much smaller footprint and lower[A1] price point. To a certain extent, we see our competition as being the legacy technologies.
The grail is the tactical navigation IMU segment, but you can think about many other applications where accelerometer MEMS on the market still hardly meet specifications.
Besides that, we will enrich our current portfolio with price-competitive MEMS sensors tailored for specific market segments, while still positioning ourselves in the high-end parts of the segments.
For example, we are currently releasing our new VS1000 product range, focused on vibration monitoring that provides a step improvement with respect to the competition in noise, resolution, non-linearity and bandwidth.
Applications such as vibration (health) monitoring in critical embedded systems (eg: train bogie monitoring) require the best of breed in MEMS. Our capacitive technology brings an incredible stability over a long period of time and environmental exposure, allowing temperature behavior compensable with simple methods.
Former limitations of bandwidth and survivability over multiple shocks have been removed, improving our position against traditional piezo technologies once more. Bandwidth is from 0 to 6kHz (-3dB), while piezo technologies are unable to address the low frequency spectrum and struggle to handle temperature variations.
Our customers are demanding: they usually do not put much trust in datasheets, as each and every application has its own profile of performance and environmental constraints. So we have close contacts with the system integrator with whom we team from time to time for system qualification.
And finally, what makes a big difference is the behavior of our sensors with time. It is one thing to make a prototype and another to manufacture large volumes with controlled quality of sensors whose performance will not deviate after 15 years of operation. If you look around, you will not find so many players that make a comparable promise. Obviously, the markets we are addressing value the quality of manufacturing and the long term stability of our product and our company.
YD: You said that you are now better positioned for the future, why is that? How do you keep a high level of innovation in your company?
PG: Being a medium size company is very challenging in the high-end sensor market, considering the breadth of competencies you have to master, and we consider it a great opportunity to be teaming with the Safran Group.
In our joint strategy with Safran, we differentiate R&T and Developments.
R&T aims at evolving the core technologies embedded in our MEMS. Our technology roadmap has four areas of focus: the die itself, the control and associated electronics, the packaging and the assembly. When you expect to reach unbeaten performance, each and every detail counts! R&T is done together with the Safran Group, because it is important for the group to propose high performance MEMS in its system roadmap. Their expertise is very valuable in areas such as electronics and packaging, as well as requirements for better system implementation and tests. Their constraints are however so specific that the electronics and packaging will not be standard at the component level.
Developments focus on assembling those technologies together to promote products in the market through our portfolio. Here Colibrys is independent: we have our own segmentation and strategy. We can readily use a die optimized for aerospace navigation and improve the noise level of the electronics to address the seismic detection business.
In terms of trends, we believe that open loop still has tremendous potential. We are far from having reached the performance ceiling, and the electronics is much simpler than closed loop. This is a fundamental point for cost, but also for the ability to have the sensor certified for safety-critical applications, like flight control systems, for example. While we are working on closed loop to bring amazing performance levels to applications that are limited today, open loop has a shiny future! To give you an order of magnitude, we are confident to reach accelerometer performance with an end-of-life bias stability of around 1 mg with open loop control. This opens up numerous markets such as flight control systems, flight display, AHRS stabilization, UAV&ROV Autopilot, north finding, directional drilling, to name just a few. Note that the “end-of-life” bias stability is a parameter that very few sensor companies dare define: take care when reading 1 mg in a data sheet – ask what the definition of the bias stability is!
Colibrys owns its state-of-the-art 6’’ wafer fab
Overall, our technology evolution addresses all areas of performance: bias stability, noise, linearity, but also environmental robustness, including temperature and (repetitive) shocks. We can relatively easily gain a factor of 10 with respect to the best-of-breed of MEMS accelerometers today. The question is not how, but for which market size?
Shock is a good example of how far we go. Where the competition measures single shocks of 5,000 or 20,000 gs, we believe that if you do not take care of it during design you can see serious degradation of performance after repetitive shocks. Unfortunately many applications, such as automotive testing and bogie monitoring for high speed trains, require resistance to repetitive shocks.
YD: Can you give us details on the development of new accelerometer products within Colibrys ?
PG: Since 2013 we have been working on the new 1000 series accelerometer. This new series uses the same, well proven, die that offers unmatched stability and combines it with open loop electronics designed with Sagem, the electronics company of the Safran Group. This electronics are low noise and provide exceptional stability over temperature. The vibration product VS1000 has already been released. The high temperature tilt TS1000T is in field testing with selected system integrators, and the inertial version is coming in 2016. We are also working on new mechanical interfaces to speed up and improve the performance of the integration of the sensor into the system.
If your question relates to a horizon after the 1000 series, we are not disclosing that. We prefer to talk about our capabilities. As I mentioned, we are far from having reached the performance ceiling. Product launches merely depend on market demand.
YD: How do you see the evolution of the accelerometer market in the coming years? What trends do you observe in the high-end segment ?
PG: The high-end will remain a niche market and the price versus volume is very different from the low-end. There is a large gap between “less than 10 USD per part” and “more than 100 USD”.
In the high-end segment, the aim is to replace legacy technologies such as electro-mechanical, quartz, and piezo that are either expensive or face limits.
One good example is the measurement while drilling (MWD) application, which is extensively penetrated by quartz inclinometers that suffer from high cost and fragility. We are working to replace the sensor with a MEMS device that is targeting precision better than +/-0.1° over lifetime with a temperature range up to 175°C and a proven mean time before failure (MTBF) under temperature and vibration of more than 1000 hours. We are currently defining that MTBF limit, it may be even better. Imagine the value for drilling operators, for whom breakage can quickly become a nightmare and a financial drain.
Another example of improved system performance is in bogie monitoring. Vibration sensors detect abnormalities in the frequency spectrum starting at low frequencies. Pushing frequency limits and reducing the noise level permits earlier detection of mechanical degradation, whether it is in the bogie structure or in the rails, which is an unmet challenge as of today.
In both cases the system is better. Either it uses a sensor whose technology enables a price cut, or its performance and robustness are improved. The value for the customer is straightforward, and the MEMS is easy to implement and is now a proven technology.
The types of applications are very diverse, but that is expertise we have at Colibrys, having invested significant resources in understanding them and developing tailored solutions.
We do not see too many competitors in that space. I believe this is mostly due to the wide range of skills and the accumulated experience of the technology platforms that it requires.
There is also something important to mention: the market has been indulgent with respect to the quality of service. Because performance was hard to find, they compromised on touchy electronic interfaces, just acceptable logistic performance, and distant support.
The bulk of the high-end segment is now becoming more demanding, not just on the performance, but also on the “package” that goes with it: on-time-delivery, small lots, on-site support, ready to use electronic interfaces. I believe that simplified system interfaces and smart sensors will probably reach the high-end segment in a 10 years period of time.
That is also an area that we are good at. We are definitely an industrial, mature, company with robust management systems and high levels of service.
YD: What can we expect from Colibrys within 5 years ?
- Improvement of the performance of our accelerometer line of business for inertial, vibration, tilt and seismic applications.
- Opening of new markets, such as drilling, structural health monitoring, flight controls and navigation,
- Proximity to the customer for sensor integration and qualification.
- Easier access to small volumes and prototypes everywhere on Earth.
YD: What are your objectives in the future? How will you sustain this growth in the future ?
PG: Our future will push us to develop sensors beyond accelerometers. Our technology and know-how can address a wide range of high performance sensors, and, if you look at it, you will find the same reduced number of high-end MEMS providers in accelerometers, gyrometers and pressure sensors. There is some hype around high-end MEMS designs, but if you look at real industrial products with a track record of sustained performance in the field, this is indeed a pretty small family of competitors! There are good reasons for that. No doubt, with the Safran Group beside us, we have now the tools in hand to be part of that exclusive family.
Our approach will remain to address the needs of the high-end aerospace and defense market, with dedicated technological platform, and derive families of catalog products to serve our traditional markets.
The potential is high, because of the fragmentation of players in field of high-end MEMS sensing and the barriers to entry. We believe that we can make a difference by bringing our industrial capabilities and development expertise together with Safran.
YD: We estimate the global IMU market will exceed $2.5B by 2018, including emerging markets like autonomous vehicles, low-orbit satellites and drones that will require small form-factor and lightweight IMUs. More generally, the industrial segment seems very promising, and such applications could represent big opportunities for you to integrate your accelerometers into these IMUs. Don’t you think it’s time to look at other markets than traditional ones? What do you think about these applications ?
PG: Technology-wise, most of the applications that you mention are already addressed by MEMS priced at around $10 per part. Their performance is just sufficient.
Now, the $2.5B opportunity for Colibrys is driven by to questions:
- Would we propose a sensor that’s 100 times better with a somewhat more expensive price tag, would there be a market for it?
- Can MEMS technologies that are today serving those markets in a cost-effective manner make a significant step in performance while staying in the same cost range?
Our answer to both questions is mostly negative. For now.
The automotive industry is currently aiming at tailoring systems around autonomous vehicles to cope with current sensor performance and provide sufficient redundancy for it to be safe enough. It is true, however, that if we put our energy into designing a very compact high-end MEMS IMU with performance and reliability comparable to navigation standard in aerospace, we can dramatically simplify the architecture and make the IMU the central part of the system, as it is already in an airplane. It is a choice of architecture. It may be that the system integrator will go in that direction. If so, we will definitely be a partner.
As for the second question, the skills and technological choices to make a highly reliable navigation-grade IMU looks to me a barrier of entry for industrial grade manufacturers. Furthermore, they are struggling with the difficulty of focusing on volume and high performance/reliability at the same time. With no doubt, the Internet Of Things will create mid- to high-volume opportunities for all classes of sensors. We will definitely be watching the choices that the different players will make in the coming ten years.
YD: Let’s talk about your foundry services. What is your specific know-how in this service? What kind of processes and materials do you offer? Who could be interested by your services ?
PG: Colibrys has 25 years’ experience of MEMS foundry. Designed and adapted for high performance devices, we adapt quality norms used in the aerospace markets to bring your device concept to a mature industrial product with high yield and low cost. The core technologies include:
- Bulk micromachining of silicon with both DRIE and wet etch
- High aspect ratio DRIE etching
- Silicon Fusion Bonding
- Thermal oxidation, LPCVD
- Metal evaporation
- Double-sided photolithography down to 1μm resolution
- Packaging and application specific test of high performance inertial and optical devices
Leveraging these capabilities, over the years Colibrys has successfully industrialized highly sophisticated products for optical telecom markets including optical switches, wavelength selective switches, variable optical attenuators, Micro-optical products for industrial applications, IR Gas detection sensors, high-speed digital printing heads based on micro magnets, individually addressable arrays of micro shutters, Diffractive Optical Elements, high energy particle detectors to name but a few.
YD: In what way are you different from competitors regarding your foundry services ?
PG: We are not just a fab.
When a customer comes to see us, we are able to understand what they want to reach, and bring the design concept to a level of maturity where the yield is high, the process is under control and the performance is guaranteed over an extended lifetime. That is what we know how to do well: we engineer concepts to industrial maturity offering custom design and solutions, test and verification on demand.
We do not simply refer to a standard library, but do custom, industrial design that starts from cooperative work at R&D level. The know-how we have developed in this area is unique.
Our engineering team has a deep understanding of the best practices in MEMS design. One fellow CEO told me he estimated that there are less than 100 good MEMS designers in the world! The Swiss cluster represents a good percentage of it, with no doubt.
We have access to a unique range of facilities that complements our range of capabilities and gives us the ability to undertake a wide range of challenges, such as:
- Expertise in high performance sensors
- Expertise in MOEMS
- Own manufacturing facility (wafer Fab)
- Local access to a cluster of micro fabrication facilities
- Established and applied quality norms and standards in our engineering and manufacturing
- Tightly controlled management system assuring optimum risk management
LATEST NEWS: Read the recent announcements presented by Patrick Gougeon at AIRTEC Congress in Munich – MEMS sensors will overpass legacy technologies…
Patrick Gougeon has 20+ years of experience in business development, organizational transformation and operational improvement in R&D intensive sectors. He joined the Safran Group in 2006 and has been appointed CEO in April 2013, after Colibrys was acquired by Sagem (Safran Group).
Source : www.colibrys.com
Related Reports and Monitors
Safran Colibrys MS1010 and MEMSIC MXA2500M High-End Accelerometers
Reverse Costing - Structural, Process & Cost Report
Sensonor STIM318 Inertial Measurement Unit (IMU)
Reverse Costing - Structural, Process & Cost Report